This application is in response to Program Announcement PA-03-127 (Integrative and Collaborative Approaches to Research) and has the long-term goal of developing algorithms for predicting the binding specificities of Cys2His2 zinc finger proteins, the most common DMA-binding domain encoded in the human genome. To do this, we propose to engineer and characterize the DMA-binding specificities of a large series of artificial Cys2His2 zinc finger proteins and to use this comprehensive dataset to develop and validate probabilistic DMA-recognition codes for Cys2His2 zinc fingers with greatly improved predictive power. The development of a large archive of engineered Cys2His2 zinc finger domains with well-defined specificities should also be generally useful to a large number of researchers interested in engineering customized "designer" zinc finger nucleases (recently shown to be powerful reagents for inducing site- specific gene targeting in human cells). Working together with Consortium Members Dr. Gary Stormo (of the Washington University School of Medicine) and Dr. Scot Wolfe (of the University of Massachusetts Medical School), we will develop methodologies, reagents, data, and algorithms relevant to understanding the DMA-binding specificities of Cys2His2 zinc fingers. These activities will enhance the long-term goals of R01-funded research programs in all three laboratories. Specifically, the proposed project will provide important zinc finger reagents and binding specificity data that will facilitate the long-term goals of R01GM069906 (Studies of NRSF/REST Zinc Finger-DMA Interactions; PI: Joung) to understand naturally occurring extended zinc finger-DMA interfaces and to design highly specific engineered zinc finger domains capable of recognizing unique addresses within mammalian genomes. Similarly, these reagents and specificity data will also facilitate the development of dimeric engineered zinc finger proteins that can bind to and regulate a single endogenous gene in eukaryotic cells (R01GM068110: Dimeric Cys2His2 Zinc Finger Proteins for Gene Targeting; PI: Wolfe). Finally, the development of algorithms to predict DMA-binding specificities of eukaryotic zinc finger transcription factors extends a fundamental goal of R01 HG000249 (DMA pattern identification and analysis; PI: Stormo) to develop computational methods to determine the specificity of DMA-binding transcription factors and identify their regulatory sites within genome sequences. [unreadable] [unreadable] [unreadable]